The Reaction of Ethylene Oxide with Biopharmaceuticals

               A recent paper by Chen and coworkers (1) attracted my attention.   These investigators reported the modification of pegylated derivative of recombinant human granulocyte colony stimulating factor (PEG-GCSF) by residual ethylene oxide in a delivery device.   The possible modification of human serum albumin (HSA) was also studied as it is an excipient for human erythropoietin (EPO) as was formulated EPO.    The modification of PEG-GCSF was studied at pH 4.0 (formulation buffer) while the modification of HSA  was studied at pH 6.7 in a ethylene oxide-sterilized plastic vials.  The products were held under stressed conditions for  PEG-GCSF (42 degrees C/54 hr) or 37 degrees C for two weeks for HSA/EPO.  Two methionine residues were modified in PEG-GCSF  while a methionine and a cysteine residue were modified in HSA in the EPO formulation.  Erythropoietin formulated with albumin was not modified by residual ethylene oxide. While there is considerable history on the reaction of ethylene oxide with proteins, the work of Chen and coworkers(1) does use mass spectrometry for the characterization of the modified biopharmaceuticals in their formulated form.

               The observation that proteins could be modified by residual ethylene oxide following sterilization is of importance as ethylene oxide is used for terminal sterilization of a variety of medical devices (2-5).   The potential of the reaction of residual ethylene oxide in sterilized device with protein was discussed in detail by Tock and Chen and 1974 (6) The emphasis of this work was aeration required to remove reactive ethylene oxide from the sterilized device.  This article contains a good discussion of the history of ethylene oxide use in the sterilization of medical devices.  Heinz Fraenkel-Conrat (7)a published the first study on the reaction of ethylene oxide with proteins.  While I don't have direct evidence , it is my sense that this particular work was part of a WWII project for modification of protein to be used as plastics.   Fraenkel-Control did find the ethylene oxide was promiscuous with modification occurs at carboxyl groups (esterification) and a variety of protein nucleophiles.  Some twenty years later, Starbuck and Busch (8) presented an extensive study on the reaction of ethylene oxide with bovine albumin or rat albumin.  The goal of this work was the preparation of Nε-hydroxyethyl derivatives which could be uses to attach chemotherapeutic agents for delivery to tumors (9).  Reaction of albumin at pH 9.0 (0.1 M Tris, 30 degrees C, 24 hr) with ethylene oxide yielded 21 different products including bis and tris-hydroxyethyl derivatives; there was also modification of methionine, tyrosine, histidine, arginine, tyrosine, and histidine was also observed.   The modification was pH dependent and consistent the need for a nucleophilic form.   Methionine was modified with considerable specificity at ph 3-4 ; there was also considerable modification of histidine.   The selective modification of methionine at acid pH  would be expected from the earlier work of Gundlach, Moore and Stein(10) and Link and Stark (11) on the reaction of methionine at mildly acidic pH with iodoacetate which, like ethylene oxide, is alkylating agent.

               The results obtained by Chen and coworkers (1) for the modification of methionine in PEG-GCSF can be rationalized by the specificity of alkylating agents for methionine at mild acid pH cited above.  Considering the work of Starbuck and Busch (8), reaction of another amino acid would be unlikely at pH4.0.  The modification of a cysteine residue in HSA with ethylene oxide is also reasonable considering the work of Starbuck and Busch (8).  The  modification of methionine is somewhat more difficult to rationalize with prior data as one would have expected the modification of a histidine residue and perhaps a lysine residues in addition a methionine residue at pH 6.7.  Starbuck and Busch (8) did use bovine albumin and rat albumin while Chen  and coworkers (1) used human serum albumin.  It is noted that Starbuck and Busch (8) do suggest that modification of the cysteine residue which is conserved in albumins (12) does predispose to subsequent modification of methionine (and lysine)   While it is unlikely, is possible that a derivative such as ethylene chlorohydrin (2-chloroethanol)(6,13) is generated during the accelerated stress conditions of Chen and coworkers (1), which could react with proteins.

Chen and coworkers (1) have reminded us of the issues such as modification of therapeutic product posed by contaminants introduced during the manufacturing of biopharmaceutical products.  The reader is encourage to read the early work by Tock and Chen (6) on the aeration of medical plastics as well as subsequent studies (14-17) on the dissipation of ethylene oxide from medical polymers and products. The reader is also encouraged to consider the possibility of generating substances such as 2-chloroethanol during the sterilization process.

a  Fraenkel-Conrat who subsequently went on to work on nucleic acids with emphasis on tobacco mosaic virus.  I became acquainted with tobacco mosaic virus in graduate school as part of a student project with Professor Milton Gordon. The project involved neutron activation analysis using the small reactor on campus.   I can't recall as whether the project was successful but a mystery novel was left in the chamber by mistake and it was sometime before the book had become "cold" enough to read

1.  Chen, L., Sloey, C., Zhang, Z., et al., Chemical modifications of therapeutic proteins induced by residual ethylene oxide, J.Pharm.Sci. 104, 731-739, 2015.

2.  Shopnick, R.I., Kazemi, M., Brettler, D.B., et al., Anaphylaxis after treatment with recombinant factor VIII,  Transfusion 36, 358-361, 1996.

3.  Lambert, B.J., Mendelson, T.A., and Craven, M.D., Radiation and ethylene oxide terminal sterilization experiences with drug eluting stent products, AAPS PharmSciTech. 12, 1116-1126, 2011.

4.  MacDonald, D., Hanzlik, J.,Sharkey, P., Parvizi, J., and Kurtz, S.M., In vivo oxidation and surface damage in retrieved ethylene oxide-sterilized total knee arthroplasties, Clin.Orthop.Relat.Res. 470, 1826-1823, 2012.

5.  Coisman, J.G., Case, J.B., Clark, N.D., Wellehan, J.F., and Ellison,G.W., Efficacy of decontamination and sterilization of a single-incision laparoscopic surgery port, Am.J.Vet.Res. 74, 934-938, 2013.

6.  Tock, R.M. and Chen, Y.C., Aeration of medical plastics, J.Biomed.Mat.Res. 8, 69-80, 1974.

7.  Fraenkel-Conrat, H., The action of 1,2-epoxides on proteins, J.Biol.Chem. 154, 227-238, 1944.

8.  Starbuck. W.C. and Busch, H., Hydroxyethylation of amino acids in plasma albumin with ethylene oxide, Biochim.Biophys.Acta 78, 594-605, 1963.

9.  Starbuck, W.C. an Busch, H., Effects of hydroxyethylation on the uptake of rat albumin by tissues of the tumor-bearing rat, Cancer Res. 22, 1206-1211, 1962.

10.  Gundlach, H.G., Moore, S., and Stein, W.H., The reaction of iodoacetate with methionine, J.Biol.Chem. 234, 1761-1764, 1959.

11. Link, T.P. and Stark, G.R., S-Methionine-29 ribonuclease A. I. Preparation and proof of structure, J.Biol.Chem. 243, 1082-1088, 1968.

12.  Lundblad, R.L., Biotechnology of the Plasma Proteins, CRC Press/Taylor and Francis, Boca Raton, Florida, 2013.

13.  Lomas, R.J., Gillan, J.L., Matthews, J.B., Ingham, F., and Kearney, J.N., An evaluation of the capacity of differently prepared demineralized bone matrices(DBM) and toxic residuals of ethylene oxide (EtOx) to provoke an inflammatory response in vitro, Biomaterials 22, 913-921, 2001)

14. Handlos, V., Kinetics of the aeration of ethylene oxide sterilized plastics, Biomaterials  1, 149-157, 1980.

15.  Vink, P. and Pleijsier, K., Aeration of ethylene glycol sterilized polymers, Biomaterials 7, 225-230, 1986.

16. Chien, Y.C., Su, P.C., Lee, L.H. and Chen, C.Y., Emission characteristics of plastic syringes sterilized with ethylene oxide -- a controlled study, J.Biomed.Mat.Res.B Appl.Biomater. 91, 579-586, 2009.

17.  Mendes, G.C., Brandao, T.R.S., and Sliva, C.L, Kinetics of ethylene oxide desorption from sterilized materials, J.AOAC Int. 96, 33-36, 2013.

Copyright, April 20, 2015 Roger L. Lundblad, Chapel Hill, North Carolina